Originally Posted by Chicken Patty
I notice that too! my i7 says on some projects 12hrs, but finish after i come back from taking a dump, ooops!
I'm glad that I'm not the only one experiencing this. I thought I must be out of my mind, because the WUs kept finishing so much faster than expected, so I took a screen shot, disabled network connections, left it running for 16 hours, and then compared results, and the estimated times were usually twice how long it actually took.
Originally Posted by FordGT90Concept
My server has point fluctations too and it runs 24/7 with relatively constant workloads:
|Statistics Date||Total Run Time (y:d:h:m:s)||Points Generated||Results Returned |
I think validation could have a lot to do with it. If you were the first to submit a result, it has to wait until someone else submits a result as well in order to validate which would then credit you points.
We also have to keep in mind that not all work units are equal. Some take longer than others and some are worth more than others. Like in my chart, it should be 8 hours run time every day but, all together, it maybe averages to 8.
I wouldn't be concerned about fluctuations because that is the nature of distrobuted computing.
Yep, that's what I was saying, my rigs are on most of the time so they almost always turn in the WU first, I have a bunch of WUs that have been pending validation for a week now...